1182
DOI 10.1002/mnfr.201400862
Mol. Nutr. Food Res. 2015, 59, 1182–1189
RESEARCH ARTICLE
Permeation of steryl ferulates through an in vitro intestinal barrier model ¨ 1 Dan Zhu1 , Davide Brambilla2 , Jean-Christophe Leroux2 and Laura Nystrom 1
Department of Health Sciences and Technology, Institute of Food, Nutrition and Health, ETH Zurich, Zurich, Switzerland 2 Department of Chemistry and Applied Biosciences, Institute of Pharmaceutical Sciences, ETH Zurich, Zurich, Switzerland Scope: Steryl ferulates (SFs) belong to the bioactive lipids contributing to the health promoting effects of whole grains. However, their intestinal absorption remains unclear. We investigated the permeation of individual SFs using an in vitro intestinal barrier model. Methods and results: An in vitro Caco-2 cell monolayer, mimicking the intestinal barrier, was used to evaluate the permeation of eight SFs from different sources. A method based on ultraperformance LC with high-resolution mass spectrometric detection was developed for their quantification. Although only a negligible amount (< 0.5%) permeated across the Caco-2 cell monolayer, some differences in the permeability coefficients were observed between individual SFs. Permeation mechanism was mainly passive diffusion. Conclusion: This work indicates that the permeation of SFs across the gut is very low. Therefore, cholesterol lowering and antioxidant activity-related health benefits of SFs most likely occur in the gut independently from absorption.
Received: November 25, 2014 Revised: January 19, 2015 Accepted: February 17, 2015
Keywords: Caco-2 / Cereal grains / Monolayer / ␥-Oryzanol / Steryl ferulate
1
Additional supporting information may be found in the online version of this article at the publisher’s web-site
Introduction
Steryl ferulates (SFs), esters of ferulic acid and phytosterols, are secondary plant metabolites present in the bran layers of grains [1]. While the mixture of SFs in rice, ␥-oryzanol, is by far the most studied, SFs have also been identified in barley, corn, rye, triticale, wheat, and wild rice [2]. SFs are known for a variety of beneficial health effects, such as cholesterol lowering activity, reduction of serum thyroid stimulating hormone levels, and attenuation of gastric hypersecretive disorder [3,4]. Additionally, SFs are recognized as antioxidants, through the hydrogen donating capacity of their ferulic acid moiety [5]. ¨ Correspondence: Prof. Dr. Laura Nystrom. Institute of Food, Nutrition and Health, ETH Zurich, Schmelzbergstrasse 9, CH-8092, Zurich, Switzerland E-mail: [email protected] Fax: +41 44 632 1123 Abbreviations: GI, gastrointestinal; HBSS, Hank’s balanced salt solutions; LY, Lucifer yellow; Papp , apparent permeability coefficient; SF, steryl ferulate; TEER, transepithelial electrical resistance; UPLC-MS, ultra-performance liquid chromatography with high-resolution mass spectrometric detection C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The sterol moieties (Fig. 1) integrated into the SFs vary between grain species and varieties, as well as environmental factors [6]. Differences in the bioactive properties of SFs have been reported earlier: for example SFs from rice (␥-oryzanol, rich in SF 1–2) had inferior antioxidant potential compared to other SFs (SF 3–6), originating from wheat bran [7]. Additionally, SFs from rye and wheat (SF 3–6) were hydrolysed more effectively than ␥-oryzanol by steryl esterase in vitro [8]. Hence, these studies reveal a different biological behavior for the individual SFs. ␥-Oryzanol has been reported to be poorly absorbed in rat [9] and rabbit [10]. Moreover, no cellular uptake of SFs was observed in human intestinal cells C2BBe1 [11]. In a recent clinical study, volunteers took a daily dose of 5% ␥oryzanol for 3 days in skimmed milk yogurt and showed that nearly 80% of intact ␥-oryzanol was recovered in the feces [12]. Nevertheless, none of these studies provide clear evidence regarding the exact intestinal absorption of individual SFs. Due to SFs’ hydrophobicity, their absorption necessitates initial release from the food matrix followed by emulsification and/or micellar solubilization by biological surfactants during digestion. The resulting oil droplets, liposomes, or micelles function as reservoir and/or transport vehicle across www.mnf-journal.com
1183
Mol. Nutr. Food Res. 2015, 59, 1182–1189
Figure 1. Chemical structures of (1) cycloartenyl ferulate; (2) 24methylenecycloartanyl ferulate; (3) campesteryl ferulate; (4) sitosteryl ferulate; (5) campestanyl ferulate; (6) sitostanyl ferulate; (7) stigmasteryl ferulate; (8) cholesteryl ferulate.
the unstirred water layer until they reach the brush border membrane of enterocytes [13]. The use of in vitro cell model monolayer provides a good starting point to investigate the permeation of SFs through the intestinal epithelium. Caco2 human colon carcinoma cells, form monolayers with tight junctions, and can be used to mimic the intestinal barrier and study food components or drugs intestinal absorption [14]. The aim of this study was to investigate the permeation of eight individual SFs (Fig. 1) through Caco-2 cell monolayers. Cycloartenyl ferulate (SF 1) is the only single SF commercially available, and is often used as model compound of individual SFs. A highly sensitive quantification method of the individual SF species based on ultra-performance liquid chromatography with high-resolution mass spectrometric detection (UPLC-MS) was developed. To the best of our knowledge, this study is the first to provide an investigation of permeability of individual SFs in a human intestinal barrier model.
2
S´anchez-Ferrer (Laboratory of Food and Soft Materials, ETH Zurich, Switzerland). Caco-2 cells were obtained from ATCC (Rockville, MD, USA). DMEM, fetal bovine serum, nonessential amino acids, Hank’s balanced salt solutions (HBSS) and 89 units/mL penicillin and 89 g/mL streptomycin were purchased from Invitrogen (Paisley, UK). Cell culture PET transwell membranes (pore size, 3.0 m, effective growth area, 0.9 cm2 ) and companion 12-well plates were from BD Biosciences (Franklin Lakes, NJ, USA). Cell Counting Kit-8 (CCK-8) was from Dojindo (Kumamoto, Japan).
2.2 Maintenance of Caco-2 cell culture Caco-2 cells were cultured in DMEM supplemented with 1% nonessential amino acids, 15% fetal bovine serum and 0.8% penicillin-streptomycin at 37⬚C in a humidified atmosphere with 5% CO2 . For maintenance, the medium was changed every second day and the cells were passaged every week. Cells of passage 49–64 were used in this study.
Materials and methods
2.1 Materials
2.3 Sample preparation
Acetic acid, acetonitrile (HPLC grade), ammonium hydroxide solution (NH4 OH), 1-butanol, BSA, hexane, isopropanol, Lucifer yellow CH dipotassium salt (LY), 1-monoolein, oleic acid, L-␣-phosphatidylcholine, and sodium taurocholate were purchased from Sigma-Aldrich (St. Louis, MO, USA). SF 1 (99%) was purchased from Wako (Osaka, Japan). Acetonitrile and isopropanol (MS grade) were purchased from Biosolve (Dieuze, France). ␥-Oryzanol was obtained from CTC Organics (Atlanta, GA, USA). Wheat bran was milled fraction from Swissmill (Zurich, Switzerland). Stigmasteryl ferulate (SF 7; 99%) and cholesteryl ferulate (SF 8; 97%) were synthesized (manuscript submitted for publication) by Dr. Antoni
2.3.1 Extraction and purification of SFs
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The extraction and purification of individual SFs were performed according to a method previously described [15]. Briefly, total lipids were extracted from wheat bran and then neutral lipids were eliminated by base-acid wash. The residue or ␥-oryzanol was separated and purified by preparative HPLC (Merck-Hitachi, Japan) using an Xbridge Prep Shield RP C18 column (Waters, Ireland) with detection at 325 nm. Mobile phase was acetonitrile/H2 O/butanol/acetic acid 88:6:4:2 at 6.6 mL/min, 25⬚C. SF 2 (96% purity as area percentage) was collected from ␥-oryzanol. SF 3 (98%) and 4 (96%) were www.mnf-journal.com
1184
D. Zhu et al.
collected from both ␥-oryzanol and wheat bran. SF 5 (95%) and 6 (98%) were from wheat bran. Quantification of SFs was performed by analytical RP-HPLC (Agilent 1100, Germany) with an XBridge Shield C18 column (Waters, Ireland), and using the same mobile phase as for preparative-HPLC with 1 mL/min flow and detection at 325 nm. SF 1 was used as an external standard for RP-HPLC quantification.
2.3.2 Preparation of SF solution The preparation method of SF in HBSS was adapted from previous studies [11, 16]. Firstly, 0.1 mM SF, 0.5 mM oleic acid, 0.25 mM 1-monoolein, and 0.3 mM phosphatidylcholine were dissolved in hexane and isopropanol (3:1, v/v). After solvent evaporation under nitrogen stream at 45⬚C, the samples were redissolved in 6.6 mM sodium taurocholate in HBSS. The mixture was sonicated (Sonorex Super RK510, Germany) in an ice bath for 1 h, and was observed to be free of precipitation. The solution was stored overnight at room temperature and centrifuged for 15 min at 3220 g. The supernatant was collected and passed through a 0.2-m PVDF syringe filter (BGB, Switzerland). For quantification in the final preparation, SF was extracted with twofold amount of hexane/isopropanol/ethyl acetate (2:1:1, v/v/v) (×2) and twofold amount of hexane/isopropanol/ethyl acetate (10:1:1, v/v/v) (×4). The organic layers were combined and redissolved in acetonitrile for quantification by RP-HPLC. SF solubilization at concentration of 70 M was achieved with the above-described method. Furthermore, the highest SF concentration in HBSS was 190 M, achieved by doubling the amount of surfactants.
2.4 Caco-2 cell viability after SF treatment Caco-2 cell viability was evaluated by a tetrazolium salt (WST8)-based colorimetric assay with the Cell Counting Kit-8. WST-8 is reduced by dehydrogenases in cells to give an orange colored formazan, directly proportional to the number of living cells. Briefly, Caco-2 cells were seeded in 96-well plates at an initial density of 5800 cells/well and grown for 2 weeks to develop a confluent monolayer. The cells were treated with 100 L of individual SF in HBSS, vehicle in HBSS (mixture of oleic acid, 1-monoolein, phosphatidylcholine, and sodium taurocholate, with the same preparation procedure without SF), as well as HBSS alone for 3 h at 37⬚C, 5% CO2 . The samples were removed and the cells were incubated with 100 L complete medium for 2 h. After this, the incubation solvent was exchanged to 10 L of CCK-8 solution and 100 L complete medium, and was incubated for another 2 h. Finally, cell viability was determined by measuring absorbance at 450 nm using a plate reader (TECAN infinite M200 Reader, Switzerland). Data are expressed as follows: viability (%) = Asample /Acontrol × 100. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Nutr. Food Res. 2015, 59, 1182–1189
2.5 Permeability experiments across Caco-2 cell monolayer 2.5.1 Quality control of Caco-2 cell monolayer To generate Caco-2 cell monolayer on the transwell membrane supports, 0.5 mL cell suspension was placed in the apical chamber at density of 180 000 cells/well, and 1.5 mL culture medium was added in the basolateral chamber. Medium was carefully changed every second day. To monitor the growth of Caco-2 monolayer, transepithelial electrical resistance (TEER) value was measured using an EVOM epithelial voltmeter with STX2 electrode (World Precision Instruments, Sarasota, FL, USA) every second day. The TEER increased with time in culture, reaching a maximum at 10–14 days, and then stabilized at 21–25 days. Only monolayers with TEER > 500 ⍀cm2 were used [14, 17].
2.5.2 Transport assay The permeability assay was performed according to a previously reported protocol [14, 17]. The medium was removed from the transwell, and cell monolayer was washed with HBSS and incubated in HBSS for 20 min. A 1.5 mL HBSS including 4% BSA (for desorption of SF from acceptor wells) was loaded in the basolateral chamber, and 0.5 mL sample solution was put in the apical side followed by immediate sampling 50 L from apical side for quantification at time 0. The sample was incubated for 3 h on a rocker (Gasser Apparatebau, Switzerland) at frequency of 46 swings/min. A 750-L sample was taken from basolateral chamber at 1 and 2 h, and replaced by the same volume of HBSS with 4% BSA. After 3 h, the total volumes from both apical and basolateral sides were collected. After transport study, Caco-2 cell monolayers were post-incubated with medium for additional 15 h. TEER was measured before every sampling during the whole experiment. To verify passage of SF through the membrane support, a control analysis of each SF was studied in the same manner without Caco-2 cell monolayer. All the collected samples were stored at –20⬚C for further quantification. The integrity of the cell monolayers was also assessed by monitoring Lucifer yellow (LY, molecular weight 521.6 Da) permeation [18]. Quantification of LY was performed using a fluorescence plate reader at Ex /Em 430/530 nm. The analyzed solutions were: (a) LY alone (1 mg/mL) at 37⬚C; (b) LY and vehicle at 37⬚C; (c) SF 1 (70 M, 190 M) with and without LY at 37⬚C and at 4⬚C; (d) eight individual SF solutions (70 M) at 37⬚C.
2.6 UPLC-MS analysis of SFs Since the concentration of SF collected from the basolateral chamber after transport assay was found to be near or below the limit of quantification of RP-HPLC with UV detector, www.mnf-journal.com
1185
Mol. Nutr. Food Res. 2015, 59, 1182–1189
UPLC-MS method based on high sensitivity and mass accuracy was developed for SF quantification. Internal standard solution was directly added to the samples collected from transport study, after which SFs were extracted with solvent as previously described (in Section 2.3.2). The separation was performed with a Waters ACQUITY UPLC system with an ACQUITY UPLC BEH C18 column (50 mm × 2.1 mm, 1.7 m particle size) using mobile phase at 0.5 mL/min, 40°C. The detection was performed with a MS system (Synapt G2) with an electrospray ionization source (negative ion mode) and a Q-TOF analyser (Waters Corp., Milford, MA, USA). Identification of SFs was according to previous studies by Mandak and Nystr¨om [2] and Fang et al. [19]. For quantification of permeated SF 8, SF 1 in acetonitrile was used as internal standard, with LOD of 1.35 ng/mL and LOQ of 4.5 ng/mL. For the quantification of other SFs, SF 8 was used as internal standard, with LOD of 1.56 ng/mL and LOQ of 5.2 ng/mL. Details of quantification are described in Supporting Information S1.
SF8 SF7 SF6
R
SF5 SF4 SF3 SF2 SF1 (190 μM) SF1 (70 μM) V (190 μM ) V (70 μM) 0
20
40
60
80
100
Viability of Caco-2 monolayer (%)
2.7 Data analysis The apparent permeability coefficient (Papp , cm/s) was calculated using the equation: Papp = ⌬Q/ (⌬t × A × C0 ) . where ⌬Q/⌬t is the rate of the compound across the monolayer, A is the surface area of the insert and C0 is the initial compound concentration in the apical chamber. All analyses were carried out at least in triplicate. Data are presented as means ± SEM. Nonparametric test with Kruskal–Wallis analysis of variance and subsequent post-hoc analysis (pairwise comparisons) was performed with IBM SPSS Statistics 19.0 (IBM Corp., Armonk, NY, USA) in order to determine significant differences between samples. The differences were considered significant for p-value < 0.05.
Figure 2. Effects of individual SFs on viability of Caco-2 monolayer cells after apical exposure (SF 2–8, 70 M; SF 1 at 70 and 190 M; vehicle (V), mixture of oleic acid, 1-monoolein, phosphatidylcholine, and sodium taurocholate, amount equals to the one in the solutions of SF at 70 and 190 M. Data are expressed as mean ± SEM (n = 3–6).
that the observed effect is due to the greater amount of vehicle used to deliver this higher concentration of SF 1 rather than SF 1 itself. Further, the cytotoxicity of the other SFs (2–8) prepared in HBSS was studied at a concentration of 70 M. Measurement of the dehydrogenase activity after incubation with individual SFs for 3 h, also did not show statistically significant cytotoxicity to the Caco-2 cell compared to the control (p > 0.05).
3.2 Effect of SF 1 on Caco-2 monolayer integrity
3
Results
3.1 Effect of steryl ferulate on Caco-2 cell monolayer viability In our study, to investigate the effect of concentration on cell viability, the cytotoxicity of steryl ferulate (SF) was studied at 70 and 190 M with the model compound cycloartenyl ferulate (SF 1) after solubilization with bile salt (sodium taurocholate) and food components (oleic acid, 1-monoolein, and phosphatidylcholine) in HBSS (Fig. 2). The cytotoxicity difference of SF 1 both at 70 and 190 M was not statistically significant compared to the control (HBSS alone) (p > 0.05). However, there seemed to be a trend that the viability of Caco2 cells was slightly impaired when increasing the SF concentration to 190 M as well as with the vehicle alone, suggesting C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
SF 1 served as a model compound to assess permeation of individual SFs. To characterize its transepithelial transport, different incubation conditions were investigated on Caco-2 monolayer. Transport at 37⬚C, but not at 4⬚C, is indicative of active-transport; while transport occurring at both temperatures suggests a passive diffusion mechanism [17, 20]. The low temperature can significantly decrease the fluidity of cell membrane and cell metabolism including the ATPases activity that is involved in the energy-dependent transport process [21, 22]. The effect of SF 1 on the tight junction of Caco-2 cell monolayer was investigated by measuring the TEER values and cotransport with fluorescent probe LY. Two concentrations of SF 1 were used, 70 M and 190 M. TEER value is temperature-dependent, thus the temperature should be well controlled [17]. In our study, TEER values were constant (around 550 ⍀•cm2 at 37⬚C, 900 ⍀•cm2 at 4⬚C) www.mnf-journal.com
D. Zhu et al.
A
600 550
TEER (
cm2)
500 450 400 350
LY SF1 (70 µM) LY + SF1 (70 µM) SF1 (190 µM) LY + SF1 (190 µM)
300 250 0
2
4
6
18
Time (h) B 1000
900
cm2)
during the incubation with SF 1 at 70 M, as well as its coincubation with LY (Fig. 3, A and B). The used vehicle had no effect on the TEER. This suggests that the integrity of Caco-2 cell monolayers was not impaired during the transport studies. When increasing the concentration to 190 M, the membrane integrity was transiently compromised, as indicated by the significant decrease of TEER values by ࣈ 25% at 37⬚C. This was likely due to the higher amount of vehicle required rather than the SFs themselves. Nevertheless, the cells regained their integrity within 18 h (Fig. 3A). Similar behavior in TEER values were also observed when performed at 4⬚C (Fig. 3B). LY is known to cross Caco-2 cell monolayer exclusively through passive paracellular diffusion [18]. When the Papp of LY is smaller than 1 × 10−6 cm/s, Caco-2 cell monolayers are suitable for the permeation study [18]. In our study, Papp of LY alone in monolayer was 5.4 × 10−8 cm/s. No significant changes for Papp of LY were observed with SF 1 at 70 M both at 4 and 37⬚C (Fig. 3C, calculation in Supporting Information S3). When the concentration of SF 1 was raised to 190 M, the Papp value of LY was slightly increased at 37⬚C, Papp was increased even further by reducing the temperature to 4⬚C, suggesting some toxic effect at high concentration of SF 1. These observations are in well agreement with cytotoxicity and TEER results; nevertheless, although the toxicity was observed, the Papp value of LY was still low ( 1 × 10−6 cm/s, whereas poorly absorbed drugs have Papp values < 1 × 10−7 cm/s [23]. The low Papp value indicates that SF 1 is a poorly absorbed compound. Moreover, SF 1 had similar Papp values at 4⬚C and 37⬚C, and no significant inhibition of permeation of SF 1 at 4⬚C, suggesting that the permeation of SF 1 is mainly through passive diffusion; nevertheless, it remains unknown whether the paracellular or transcellular diffusion occurs in the Caco-2 cell monolayer.
Mol. Nutr. Food Res. 2015, 59, 1182–1189
TEER (
1186
800
700
600 LY + SF1 (70 µM) LY + SF1 (190 µM)
500 0
1
2
3
4
5
6
Time (h)
C LY +SF1 (190 µM) at 4°C
*
LY + SF1 (190 µM) at 37°C
LY + SF1 (70 µM) at 4°C
* LY + SF1 (70 µM) at 37°C
LY + V (70 µM )
LY
0
4
8
12
16
Permeability coefficient of LY (×10 - 8 cm/s)
3.3 Individual SFs The ability of individual SFs to permeate was investigated using Caco-2 cell monolayer. Given that 190 M of SFs showed some cytotoxicity, likely due to the higher amount of vehicle, all additional SFs were tested at 70 M. Monitoring the integrity of the monolayer (through TEER measurements) suggests no significant cytotoxicity or compromise of the C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 3. Effects of apical exposure of cycloartenyl ferulate (SF 1) in different conditions on the tight junction of Caco-2 monolayer: (A) Variation of TEER of Caco-2 monolayer at 37⬚C. (B) Variation of TEER of Caco-2 monolayer at 4⬚C. (C) Permeability coefficient of LY in Caco-2 monolayer. Data are expressed as mean ± SEM (n = 3–6). An asterisk indicates a significant difference between groups (p < 0.05).
www.mnf-journal.com
1187
Mol. Nutr. Food Res. 2015, 59, 1182–1189
A SF1 (70 µM) at 37°C SF1 (70 µM) + LY at 37°C SF1 (70 µM) + LY at 4°C SF1 (190 µM) at 37°C SF1 (190 µM) + LY at 37°C SF1 (190 µM) + LY at 4°C
0.3
0.2
0.1
0.0 0
SF1 SF2 SF3 SF4 SF5 SF6 SF7 SF8
0.3
1
2
3
Time (h)
Cumulative percentage transported (%)
Cumulative percentage transported (%)
A
0.2
0.1
0.0 0
B
1
2
3
Time (h)
B
SF1 (190 µM) + LY at 4°C
SF8
SF1 (190 µM) + LY at 37°C
SF7 SF1 (190 µM) at 37°C
SF6 SF1 (70 µM) + LY at 4°C
SF5 SF1 (70 µM) + LY at 37°C
SF4
SF1 (70 µM) at 37°C
SF3
0
4
8
Permeability coefficient (×10
12 -8
SF2
16
cm/s)
Figure 4. Permeation of cycloartenyl ferulate (SF 1) in different conditions in Caco-2 monolayer: (A) transport percentage of SF 1 (solid line for 70 M, dashed line for 190 M) over time; (B) comparison of the permeability coefficients of SF 1. Data are expressed as mean ± SEM (n = 6–9).
monolayer, being the resistance consistently at 550 ⍀•cm2 (Supporting Information S5). Although their transport ratios were very low, all the individual SFs showed a time-dependent passage across the model (Fig. 5A). Only 0.23% of SF 1 and 0.22% of SF 6 permeated to the receiver chamber, while for SF 8, the permeation ratio was only 0.04%. In this study, mass recovery was consistently around 100% for all individual SFs, which indicates no significant metabolism or degradation of SFs in monolayer. Generally, all of these SFs had very low Papp value (1.69 × 10−8 –1.10 × 10−7 cm/s), suggesting they are poorly absorbed compounds (Fig. 5B; calculation in Supporting Information S5); nevertheless, up to a 6.5-fold difference was observed between the highest and lowest Papp values (SF 1 versus 8). Statistically, Papp of SF 1 was significantly higher than SF 7 and 8, as well as SF 5 was significantly higher than 8. This study confirmed that individual SFs have very low permeation, while still having significant differences in their permeation values between SF species. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
SF1
0
4
8
12
16
Permeability coefficient (×10 - 8 cm/s)
Figure 5. Permeation of individual SFs (SF 1–8, 70 M) in Caco2 monolayer at 37⬚C: (A) transport percentage of SFs over time; (B) comparison of the permeability coefficients of SFs. Data are expressed as mean ± SEM (n = 6–9). An asterisk indicates a significant difference between groups (p < 0.05).
4
Discussion
Previous in vivo studies suggested the low absorption behavior of SFs. Fujiwara and coworkers reported that after oral administration of radiolabeled SF to rats, only a small fraction of radioactivity transferred to blood, with a peak concentration corresponding to 0.06% of the dose 4 h after administration [9]. From the available clinical human study by Lubinus and coworkers, no intact SF was reported in the blood sample after consumption of SFs enriched yogurt, and only cycloartenol, which was thought as the product from SF 1, was significantly increased by 23.7% compared on day 3 and day 0 [12]. This suggested that metabolism or degradation of SF 1 might have occurred in the gastrointestinal (GI) tract. Furthermore, Umehara and coworkers reported that in healthy volunteers the peak plasma concentration of SFs was 21–107 ng/mL www.mnf-journal.com
1188
D. Zhu et al.
(equivalent to 0.01–0.05% of the total administered dose) after an oral dose of 600 mg, and 112 ng/mL (< 0.3% of the dose) after repetitive oral dose of 100 mg three times a day for 10 days [24]. However, no information was so far available about the intestinal absorption of individual SFs, hence a different behavior between individual SFs cannot be excluded. In this study, we optimized the permeation conditions using a model compound SF 1, to investigate the permeation of an array of naturally occurring SFs. Although very small amount (less than 0.5%) of individual SFs permeated across the Caco-2 cell monolayer, we still observed that some differences of Papp were found between individual SFs, Papp of SF 1 > 7, 1 > 8, and 3 > 7. Examining the structural differences between these pairs of compounds lead us to believe that SF structures may influence their ability to permeate across the monolayer. SF 1 and 7 or 8 differ in both their sterol rings and side chains, SF 3 and 7 only differ in the presence and absence of one carbon at C-24 position. There is some evidence within literature that various sterols have different absorptions in the animal model, which was associated to the variation in their side chain (campesterol > sitosterol > stigmasterol) and saturation of double bond (sitosterol > sitostanol) [25]. In the view of physicochemical properties of eight SFs, we applied semi-empirical quantum mechanics method (AM1) for molecular simulation to explain their Papp differences. The dipole moment and quantitative structureactivity relationship parameters (logP, surface area, volume, hydration energy, etc.) were obtained from the quantum mechanics simulations. However, we could not identify a distinct trend between their physicochemical properties and permeability (Supporting Information S6). Moreover, high mass recovery after permeation in this study suggested there was no significant hydrolysis, metabolism, or degradation of SF by the cells. SFs were initially hypothesized to be hydrolysed to free ferulic acid and phytosterol, and could subsequently be absorbed [26]. Indeed, some in vitro studies have shown that SFs are substrate for cholesteryl esterase from animal pancreas and microorganism, and cholesteryl esterase has strong preference for desmethylsteryl ferulate [8, 27, 28]. However, Huang found that the decrease in SFs concentration was not accompanied by either an increase in free phytosterols or in ferulic acid in in vitro cell model, indicating that hydrolysis of SF is still unclear [11]. Moreover, the hydrolysis is still possibly present by pancreatic and microbial cholesteryl esterase in vivo [12]. The Caco-2 cell model lacks gut microorganisms that may secrete cholesterol esterase; the significant hydrolysis of SF is not expected. Without additional analysis we cannot entirely exclude the possibility of SF degradation or hydrolysis occurring to a minor degree. Furthermore, the similar permeation of SF 1 both at 4⬚C and 37⬚C indicated that the transport mechanism functions primarily through passive diffusion. Nevertheless, it remains to be determined whether it occurs through paracellular or transcellular diffusion. The similar behaviors in permeation between SFs (SF 2–8) to
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Nutr. Food Res. 2015, 59, 1182–1189
SF 1 (e.g. the similar TEER recording or low Papp values) lead us to conclude that they also permeate through an analogous passive diffusion mechanism. Still, active transport cannot be fully excluded within this study. In our study, the low Papp values resulted in extremely low levels of transported SFs and therefore in an analytical limitation. This hinders further analysis of potential hydrolytic products of SFs, as well as the analysis of a possible active transport. SFs are primarily found in the bran of certain cereal grains and seeds. Their biosynthetic pathway as well as location within the plant cells remains unknown. Their function in plants may be related to protection against environmental stress [29], antioxidant property [30], and regulation of microbial activity in the grains [1]. Dietary intake of SFs usually comes from cereal grains like rice, wheat and corn. However, according to our previous study, the bioaccessibility of SFs was found to be almost negligible (0.1%) from a grain matrix [2], which indicates that SFs do not need to be highly absorbable from the complex grain matrix to provide their functional effect. Furthermore, the low absorption behavior, confirmed in this study, suggests that the cholesterol-lowering and antioxidant activity-related health benefits associated with SFs most likely occur in the gut independently from absorption. For instance, a recent work by M¨akynen and coworkers showed that SFs (␥-oryzanol) significantly inhibited incorporation of cholesterol into synthetic micelles, and also SFs had ability to significantly decrease the apical uptake of cholesterol into Caco-2 cells [31]. Moreover, many of the poorly absorbed dietary antioxidants indeed play an important role in protecting the GI tract from oxidative damage without necessarily being absorbed [32]. Their concentration can be much higher in the lumen of GI tract than are ever achieved in plasma or tissues, and their antioxidant action in the GI tract can delay the development of inflammation and related disease (e.g. cancer) in stomach or colon [33]. SFs showed the potential to ameliorate colonic inflammation in colitis mice model by inhibiting NF-B activity and decreasing proinflammatory cytokines and COX-2 levels due to their antioxidant effects [34]. Furthermore, they may act as antioxidants against reactive oxygen species produced by intestinal microflora. Little is known about the bioactivities of SFs in the digesta, an area that remains to be explored. The results of this work confirm that all the individual SFs share the same behavior: they are poorly absorbed compounds and their health benefits most likely occur in the gut independently from absorption. This work was supported by ETH Zurich and doctoral scholarship from Chinese Scholarship Council (CSC). D. Brambilla gratefully acknowledges support from the ETH Zurich Postdoctoral Fellowship (2012-01). We thank Dr. Antoni S´anchez-Ferrer (Laboratory of Food and Soft Materials, ETH Zurich, Switzerland) for providing stigmasteryl and cholesteryl ferulate synthesis and computer simulation, and thank Dr. Anja Rahn for English editing. The authors have declared no conflict of interest.
www.mnf-journal.com
Mol. Nutr. Food Res. 2015, 59, 1182–1189
5
References
[1] Seitz, L. M., Stanol and sterol esters of ferulic and p-coumaric acids in wheat, corn, rye, and triticale. J. Agric. Food Chem. 1989, 37, 662–667. ¨ [2] Mandak, E., Nystrom, L., The effect of in vitro digestion on steryl ferulates from rice (Oryza sativa L.) and other grains. J. Agric. Food Chem. 2012, 60, 6123–6130.
1189 [18] Yu, H., Huang, Q., Investigation of the absorption mechanism of solubilized curcumin using Caco-2 cell monolayers. J. Agric. Food Chem. 2011, 59, 9120–9126. [19] Fang, N., Yu, S., Badger, T. M., Characterization of triterpene alcohol and sterol ferulates in rice bran using LC-MS/MS. J. Agric. Food Chem. 2003, 51, 3260–3267.
[3] Ghatak, S. B., Panchal, S. J., Gamma-oryzanol-A multipurpose steryl ferulate. Curr. Nutr. Food Sci. 2011, 7, 10–20.
´ [20] Rieux, A. d., Ragnarsson, E. G. E., Gullberg, E., Preat, V. et al., Transport of nanoparticles across an in vitro model of the human intestinal follicle associated epithelium. Eur. J. Pharm. Sci. 2005, 25, 455–465.
[4] Cicero, A., Derosa, G., Rice bran and its main components: potential role in the management of coronary risk factors. Curr. Top. Nutraceutical Res. 2005, 3, 29–46.
[21] Los, D. A., Murata, N., Membrane fluidity and its roles in the perception of environmental signals. Biochim. Biophys. Acta 2004, 1666, 142–157.
¨ [5] Nystrom, L., in: Xu, Z., Howard, L. R. (Eds.), Analysis of Antioxidant-Rich Phytochemicals, Wiley-Blackwell, Oxford, UK 2012, pp. 313–351.
[22] Holtzman, E., in: Holtzman, E. (Ed.), Lysosomes, Springer US 1989, pp. 25–92.
[6] Bergman, C. J., Xu, Z., Genotype and environment effects on tocopherol, tocotrienol, and ␥-oryzanol contents of southern U.S. rice. Cereal Chem. 2003, 80, 446–449. ¨ ¨ [7] Nystrom, L., Makinen, M., Lampi, A. M., Piironen, V., Antioxidant activity of steryl ferulate extracts from rye and wheat bran. J. Agric. Food Chem. 2005, 53, 2503–2510. ¨ [8] Nystrom, L., Moreau, R. A., Lampi, A.-M., Hicks, K. B., Piironen, V., Enzymatic hydrolysis of steryl ferulates and steryl glycosides. Eur. Food Res. Technol. 2008, 227, 727–733. [9] Fujiwara, S., Sakurai, S., Sugimoto, I., Awata, N., Absorption and metabolism of gamma-oryzanol in rats. Chem. Pharm. Bull. 1983, 31, 645–652. [10] Fujiwara, S., Sakurai, S., Noumi, K., Sugimoto, I., Awata, N., Metabolism of gamma-oryzanol in rabbit. Yakugaku zasshi 1980, 100, 1011–1018. [11] Huang, C. J., Potential functionality and digestibility of oryzanol as determined using in vitro cell culture models. PhD Thesis. Louisiana State University 2003. [12] Lubinus, T., Barnsteiner, A., Skurk, T., Hauner, H., Engel, K. H., Fate of dietary phytosteryl/-stanyl esters: analysis of individual intact esters in human feces. Eur. J. Nutr. 2013, 52, 997–1013. [13] Carey, M. C., Small, D. M., Bliss, C. M., Lipid digestion and absorption. Annu. Rev. Physiol. 1983, 45, 651–677. [14] Hubatsch, I., Ragnarsson, E. G., Artursson, P., Determination of drug permeability and prediction of drug absorption in Caco-2 monolayers. Nat. Protoc. 2007, 2, 2111–2119. ¨ [15] Mandak, E., Zhu, D., Godany, T. A., Nystrom, L., Fourier transform infrared spectroscopy and Raman spectroscopy as tools for identification of steryl ferulates. J. Agric. Food Chem. 2013, 61, 2446–2452. [16] Hamada, T., Goto, H., Yamahira, T., Sugawara, T. et al., Solubility in and affinity for the bile salt micelle of plant sterols are important determinants of their intestinal absorption in rats. Lipids 2006, 41, 551–556. ˚ o, ¨ J., Taipalensuu, J., Ocklind, G., Artursson, [17] Tavelin, S., Grasj P., in: Wise, C. (Ed.), Epithelial Cell Culture Protocols, Humana Press, Totowa 2002, pp. 233–272.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[23] Artursson, P., Palm, K., Luthman, K., Caco-2 monolayers in experimental and theoretical predictions of drug transport. Adv. Drug Deliv. Rev. 2001, 46, 27–43. [24] Umehara, K., Shimokawa, Y., Miyamoto, G., Effect of gamma-oryzanol on cytochrome P450 activities in human liver microsomes. Biol. Pharm. Bull. 2004, 27, 1151–1153. [25] Ling, W. H., Jones, P. J., Dietary phytosterols: a review of metabolism, benefits and side effects. Life Sci. 1995, 57, 195– 206. [26] Berger, A., Rein, D., Schafer, A., Monnard, I. et al., Similar cholesterol-lowering properties of rice bran oil, with varied gamma-oryzanol, in mildly hypercholesterolemic men. Eur. J. Nutr. 2005, 44, 163–173. [27] Miller, A., Majauskaite, L., Engel, K.-H., Enzyme-catalyzed hydrolysis of ␥-oryzanol. Eur. Food Res. Technol. 2004, 218, 349–354. [28] Moreau, R. A., Hicks, K. B., The in vitro hydrolysis of phytosterol conjugates in food matrices by mammalian digestive enzymes. Lipids 2004, 39, 769–776. [29] Britz, S. J., Prasad, P. V., Moreau, R. A., Allen, L. H. J. et al., Influence of growth temperature on the amounts of tocopherols, tocotrienols, and gamma-oryzanol in brown rice. J. Agric. Food Chem. 2007, 55, 7559–7565. [30] Nantiyakul, N., Furse, S., Fisk, I., Foster, T. et al., Phytochemical composition of Oryza sativa (rice) bran oil bodies in crude and purified Isolates. J. Amer. Oil Chem. Soc. 2012, 89, 1867– 1872. ¨ [31] Makynen, K., Chitchumroonchokchai, C., Adisakwattana, S., Failla, M., Ariyapitipun, T., Effect of gamma-oryzanol on the bioaccessibility and synthesis of cholesterol. Eur. Rev. Med. Pharmacol. Sci. 2012, 16, 49–56. [32] Stahl, W., vanden Berg, H., Arthur, J., Bast, A. et al., Bioavailability and metabolism. Mol. Aspects Med. 2002, 23, 39–100. [33] Halliwell, B., Zhao, K., Whiteman, M., The gastrointestinal tract: a major site of antioxidant action? Free Radic. Res. 2000, 33, 819–830. [34] Islam, M. S., Murata, T., Fujisawa, M., Nagasaka, R. et al., Anti-inflammatory effects of phytosteryl ferulates in colitis induced by dextran sulphate sodium in mice. Br. J. Pharmacol. 2008, 154, 812–824.
www.mnf-journal.com
DOI 10.1002/mnfr.201400862
Mol. Nutr. Food Res. 2015, 59, 1182–1189
RESEARCH ARTICLE
Permeation of steryl ferulates through an in vitro intestinal barrier model ¨ 1 Dan Zhu1 , Davide Brambilla2 , Jean-Christophe Leroux2 and Laura Nystrom 1
Department of Health Sciences and Technology, Institute of Food, Nutrition and Health, ETH Zurich, Zurich, Switzerland 2 Department of Chemistry and Applied Biosciences, Institute of Pharmaceutical Sciences, ETH Zurich, Zurich, Switzerland Scope: Steryl ferulates (SFs) belong to the bioactive lipids contributing to the health promoting effects of whole grains. However, their intestinal absorption remains unclear. We investigated the permeation of individual SFs using an in vitro intestinal barrier model. Methods and results: An in vitro Caco-2 cell monolayer, mimicking the intestinal barrier, was used to evaluate the permeation of eight SFs from different sources. A method based on ultraperformance LC with high-resolution mass spectrometric detection was developed for their quantification. Although only a negligible amount (< 0.5%) permeated across the Caco-2 cell monolayer, some differences in the permeability coefficients were observed between individual SFs. Permeation mechanism was mainly passive diffusion. Conclusion: This work indicates that the permeation of SFs across the gut is very low. Therefore, cholesterol lowering and antioxidant activity-related health benefits of SFs most likely occur in the gut independently from absorption.
Received: November 25, 2014 Revised: January 19, 2015 Accepted: February 17, 2015
Keywords: Caco-2 / Cereal grains / Monolayer / ␥-Oryzanol / Steryl ferulate
1
Additional supporting information may be found in the online version of this article at the publisher’s web-site
Introduction
Steryl ferulates (SFs), esters of ferulic acid and phytosterols, are secondary plant metabolites present in the bran layers of grains [1]. While the mixture of SFs in rice, ␥-oryzanol, is by far the most studied, SFs have also been identified in barley, corn, rye, triticale, wheat, and wild rice [2]. SFs are known for a variety of beneficial health effects, such as cholesterol lowering activity, reduction of serum thyroid stimulating hormone levels, and attenuation of gastric hypersecretive disorder [3,4]. Additionally, SFs are recognized as antioxidants, through the hydrogen donating capacity of their ferulic acid moiety [5]. ¨ Correspondence: Prof. Dr. Laura Nystrom. Institute of Food, Nutrition and Health, ETH Zurich, Schmelzbergstrasse 9, CH-8092, Zurich, Switzerland E-mail: [email protected] Fax: +41 44 632 1123 Abbreviations: GI, gastrointestinal; HBSS, Hank’s balanced salt solutions; LY, Lucifer yellow; Papp , apparent permeability coefficient; SF, steryl ferulate; TEER, transepithelial electrical resistance; UPLC-MS, ultra-performance liquid chromatography with high-resolution mass spectrometric detection C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The sterol moieties (Fig. 1) integrated into the SFs vary between grain species and varieties, as well as environmental factors [6]. Differences in the bioactive properties of SFs have been reported earlier: for example SFs from rice (␥-oryzanol, rich in SF 1–2) had inferior antioxidant potential compared to other SFs (SF 3–6), originating from wheat bran [7]. Additionally, SFs from rye and wheat (SF 3–6) were hydrolysed more effectively than ␥-oryzanol by steryl esterase in vitro [8]. Hence, these studies reveal a different biological behavior for the individual SFs. ␥-Oryzanol has been reported to be poorly absorbed in rat [9] and rabbit [10]. Moreover, no cellular uptake of SFs was observed in human intestinal cells C2BBe1 [11]. In a recent clinical study, volunteers took a daily dose of 5% ␥oryzanol for 3 days in skimmed milk yogurt and showed that nearly 80% of intact ␥-oryzanol was recovered in the feces [12]. Nevertheless, none of these studies provide clear evidence regarding the exact intestinal absorption of individual SFs. Due to SFs’ hydrophobicity, their absorption necessitates initial release from the food matrix followed by emulsification and/or micellar solubilization by biological surfactants during digestion. The resulting oil droplets, liposomes, or micelles function as reservoir and/or transport vehicle across www.mnf-journal.com
1183
Mol. Nutr. Food Res. 2015, 59, 1182–1189
Figure 1. Chemical structures of (1) cycloartenyl ferulate; (2) 24methylenecycloartanyl ferulate; (3) campesteryl ferulate; (4) sitosteryl ferulate; (5) campestanyl ferulate; (6) sitostanyl ferulate; (7) stigmasteryl ferulate; (8) cholesteryl ferulate.
the unstirred water layer until they reach the brush border membrane of enterocytes [13]. The use of in vitro cell model monolayer provides a good starting point to investigate the permeation of SFs through the intestinal epithelium. Caco2 human colon carcinoma cells, form monolayers with tight junctions, and can be used to mimic the intestinal barrier and study food components or drugs intestinal absorption [14]. The aim of this study was to investigate the permeation of eight individual SFs (Fig. 1) through Caco-2 cell monolayers. Cycloartenyl ferulate (SF 1) is the only single SF commercially available, and is often used as model compound of individual SFs. A highly sensitive quantification method of the individual SF species based on ultra-performance liquid chromatography with high-resolution mass spectrometric detection (UPLC-MS) was developed. To the best of our knowledge, this study is the first to provide an investigation of permeability of individual SFs in a human intestinal barrier model.
2
S´anchez-Ferrer (Laboratory of Food and Soft Materials, ETH Zurich, Switzerland). Caco-2 cells were obtained from ATCC (Rockville, MD, USA). DMEM, fetal bovine serum, nonessential amino acids, Hank’s balanced salt solutions (HBSS) and 89 units/mL penicillin and 89 g/mL streptomycin were purchased from Invitrogen (Paisley, UK). Cell culture PET transwell membranes (pore size, 3.0 m, effective growth area, 0.9 cm2 ) and companion 12-well plates were from BD Biosciences (Franklin Lakes, NJ, USA). Cell Counting Kit-8 (CCK-8) was from Dojindo (Kumamoto, Japan).
2.2 Maintenance of Caco-2 cell culture Caco-2 cells were cultured in DMEM supplemented with 1% nonessential amino acids, 15% fetal bovine serum and 0.8% penicillin-streptomycin at 37⬚C in a humidified atmosphere with 5% CO2 . For maintenance, the medium was changed every second day and the cells were passaged every week. Cells of passage 49–64 were used in this study.
Materials and methods
2.1 Materials
2.3 Sample preparation
Acetic acid, acetonitrile (HPLC grade), ammonium hydroxide solution (NH4 OH), 1-butanol, BSA, hexane, isopropanol, Lucifer yellow CH dipotassium salt (LY), 1-monoolein, oleic acid, L-␣-phosphatidylcholine, and sodium taurocholate were purchased from Sigma-Aldrich (St. Louis, MO, USA). SF 1 (99%) was purchased from Wako (Osaka, Japan). Acetonitrile and isopropanol (MS grade) were purchased from Biosolve (Dieuze, France). ␥-Oryzanol was obtained from CTC Organics (Atlanta, GA, USA). Wheat bran was milled fraction from Swissmill (Zurich, Switzerland). Stigmasteryl ferulate (SF 7; 99%) and cholesteryl ferulate (SF 8; 97%) were synthesized (manuscript submitted for publication) by Dr. Antoni
2.3.1 Extraction and purification of SFs
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
The extraction and purification of individual SFs were performed according to a method previously described [15]. Briefly, total lipids were extracted from wheat bran and then neutral lipids were eliminated by base-acid wash. The residue or ␥-oryzanol was separated and purified by preparative HPLC (Merck-Hitachi, Japan) using an Xbridge Prep Shield RP C18 column (Waters, Ireland) with detection at 325 nm. Mobile phase was acetonitrile/H2 O/butanol/acetic acid 88:6:4:2 at 6.6 mL/min, 25⬚C. SF 2 (96% purity as area percentage) was collected from ␥-oryzanol. SF 3 (98%) and 4 (96%) were www.mnf-journal.com
1184
D. Zhu et al.
collected from both ␥-oryzanol and wheat bran. SF 5 (95%) and 6 (98%) were from wheat bran. Quantification of SFs was performed by analytical RP-HPLC (Agilent 1100, Germany) with an XBridge Shield C18 column (Waters, Ireland), and using the same mobile phase as for preparative-HPLC with 1 mL/min flow and detection at 325 nm. SF 1 was used as an external standard for RP-HPLC quantification.
2.3.2 Preparation of SF solution The preparation method of SF in HBSS was adapted from previous studies [11, 16]. Firstly, 0.1 mM SF, 0.5 mM oleic acid, 0.25 mM 1-monoolein, and 0.3 mM phosphatidylcholine were dissolved in hexane and isopropanol (3:1, v/v). After solvent evaporation under nitrogen stream at 45⬚C, the samples were redissolved in 6.6 mM sodium taurocholate in HBSS. The mixture was sonicated (Sonorex Super RK510, Germany) in an ice bath for 1 h, and was observed to be free of precipitation. The solution was stored overnight at room temperature and centrifuged for 15 min at 3220 g. The supernatant was collected and passed through a 0.2-m PVDF syringe filter (BGB, Switzerland). For quantification in the final preparation, SF was extracted with twofold amount of hexane/isopropanol/ethyl acetate (2:1:1, v/v/v) (×2) and twofold amount of hexane/isopropanol/ethyl acetate (10:1:1, v/v/v) (×4). The organic layers were combined and redissolved in acetonitrile for quantification by RP-HPLC. SF solubilization at concentration of 70 M was achieved with the above-described method. Furthermore, the highest SF concentration in HBSS was 190 M, achieved by doubling the amount of surfactants.
2.4 Caco-2 cell viability after SF treatment Caco-2 cell viability was evaluated by a tetrazolium salt (WST8)-based colorimetric assay with the Cell Counting Kit-8. WST-8 is reduced by dehydrogenases in cells to give an orange colored formazan, directly proportional to the number of living cells. Briefly, Caco-2 cells were seeded in 96-well plates at an initial density of 5800 cells/well and grown for 2 weeks to develop a confluent monolayer. The cells were treated with 100 L of individual SF in HBSS, vehicle in HBSS (mixture of oleic acid, 1-monoolein, phosphatidylcholine, and sodium taurocholate, with the same preparation procedure without SF), as well as HBSS alone for 3 h at 37⬚C, 5% CO2 . The samples were removed and the cells were incubated with 100 L complete medium for 2 h. After this, the incubation solvent was exchanged to 10 L of CCK-8 solution and 100 L complete medium, and was incubated for another 2 h. Finally, cell viability was determined by measuring absorbance at 450 nm using a plate reader (TECAN infinite M200 Reader, Switzerland). Data are expressed as follows: viability (%) = Asample /Acontrol × 100. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Nutr. Food Res. 2015, 59, 1182–1189
2.5 Permeability experiments across Caco-2 cell monolayer 2.5.1 Quality control of Caco-2 cell monolayer To generate Caco-2 cell monolayer on the transwell membrane supports, 0.5 mL cell suspension was placed in the apical chamber at density of 180 000 cells/well, and 1.5 mL culture medium was added in the basolateral chamber. Medium was carefully changed every second day. To monitor the growth of Caco-2 monolayer, transepithelial electrical resistance (TEER) value was measured using an EVOM epithelial voltmeter with STX2 electrode (World Precision Instruments, Sarasota, FL, USA) every second day. The TEER increased with time in culture, reaching a maximum at 10–14 days, and then stabilized at 21–25 days. Only monolayers with TEER > 500 ⍀cm2 were used [14, 17].
2.5.2 Transport assay The permeability assay was performed according to a previously reported protocol [14, 17]. The medium was removed from the transwell, and cell monolayer was washed with HBSS and incubated in HBSS for 20 min. A 1.5 mL HBSS including 4% BSA (for desorption of SF from acceptor wells) was loaded in the basolateral chamber, and 0.5 mL sample solution was put in the apical side followed by immediate sampling 50 L from apical side for quantification at time 0. The sample was incubated for 3 h on a rocker (Gasser Apparatebau, Switzerland) at frequency of 46 swings/min. A 750-L sample was taken from basolateral chamber at 1 and 2 h, and replaced by the same volume of HBSS with 4% BSA. After 3 h, the total volumes from both apical and basolateral sides were collected. After transport study, Caco-2 cell monolayers were post-incubated with medium for additional 15 h. TEER was measured before every sampling during the whole experiment. To verify passage of SF through the membrane support, a control analysis of each SF was studied in the same manner without Caco-2 cell monolayer. All the collected samples were stored at –20⬚C for further quantification. The integrity of the cell monolayers was also assessed by monitoring Lucifer yellow (LY, molecular weight 521.6 Da) permeation [18]. Quantification of LY was performed using a fluorescence plate reader at Ex /Em 430/530 nm. The analyzed solutions were: (a) LY alone (1 mg/mL) at 37⬚C; (b) LY and vehicle at 37⬚C; (c) SF 1 (70 M, 190 M) with and without LY at 37⬚C and at 4⬚C; (d) eight individual SF solutions (70 M) at 37⬚C.
2.6 UPLC-MS analysis of SFs Since the concentration of SF collected from the basolateral chamber after transport assay was found to be near or below the limit of quantification of RP-HPLC with UV detector, www.mnf-journal.com
1185
Mol. Nutr. Food Res. 2015, 59, 1182–1189
UPLC-MS method based on high sensitivity and mass accuracy was developed for SF quantification. Internal standard solution was directly added to the samples collected from transport study, after which SFs were extracted with solvent as previously described (in Section 2.3.2). The separation was performed with a Waters ACQUITY UPLC system with an ACQUITY UPLC BEH C18 column (50 mm × 2.1 mm, 1.7 m particle size) using mobile phase at 0.5 mL/min, 40°C. The detection was performed with a MS system (Synapt G2) with an electrospray ionization source (negative ion mode) and a Q-TOF analyser (Waters Corp., Milford, MA, USA). Identification of SFs was according to previous studies by Mandak and Nystr¨om [2] and Fang et al. [19]. For quantification of permeated SF 8, SF 1 in acetonitrile was used as internal standard, with LOD of 1.35 ng/mL and LOQ of 4.5 ng/mL. For the quantification of other SFs, SF 8 was used as internal standard, with LOD of 1.56 ng/mL and LOQ of 5.2 ng/mL. Details of quantification are described in Supporting Information S1.
SF8 SF7 SF6
R
SF5 SF4 SF3 SF2 SF1 (190 μM) SF1 (70 μM) V (190 μM ) V (70 μM) 0
20
40
60
80
100
Viability of Caco-2 monolayer (%)
2.7 Data analysis The apparent permeability coefficient (Papp , cm/s) was calculated using the equation: Papp = ⌬Q/ (⌬t × A × C0 ) . where ⌬Q/⌬t is the rate of the compound across the monolayer, A is the surface area of the insert and C0 is the initial compound concentration in the apical chamber. All analyses were carried out at least in triplicate. Data are presented as means ± SEM. Nonparametric test with Kruskal–Wallis analysis of variance and subsequent post-hoc analysis (pairwise comparisons) was performed with IBM SPSS Statistics 19.0 (IBM Corp., Armonk, NY, USA) in order to determine significant differences between samples. The differences were considered significant for p-value < 0.05.
Figure 2. Effects of individual SFs on viability of Caco-2 monolayer cells after apical exposure (SF 2–8, 70 M; SF 1 at 70 and 190 M; vehicle (V), mixture of oleic acid, 1-monoolein, phosphatidylcholine, and sodium taurocholate, amount equals to the one in the solutions of SF at 70 and 190 M. Data are expressed as mean ± SEM (n = 3–6).
that the observed effect is due to the greater amount of vehicle used to deliver this higher concentration of SF 1 rather than SF 1 itself. Further, the cytotoxicity of the other SFs (2–8) prepared in HBSS was studied at a concentration of 70 M. Measurement of the dehydrogenase activity after incubation with individual SFs for 3 h, also did not show statistically significant cytotoxicity to the Caco-2 cell compared to the control (p > 0.05).
3.2 Effect of SF 1 on Caco-2 monolayer integrity
3
Results
3.1 Effect of steryl ferulate on Caco-2 cell monolayer viability In our study, to investigate the effect of concentration on cell viability, the cytotoxicity of steryl ferulate (SF) was studied at 70 and 190 M with the model compound cycloartenyl ferulate (SF 1) after solubilization with bile salt (sodium taurocholate) and food components (oleic acid, 1-monoolein, and phosphatidylcholine) in HBSS (Fig. 2). The cytotoxicity difference of SF 1 both at 70 and 190 M was not statistically significant compared to the control (HBSS alone) (p > 0.05). However, there seemed to be a trend that the viability of Caco2 cells was slightly impaired when increasing the SF concentration to 190 M as well as with the vehicle alone, suggesting C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
SF 1 served as a model compound to assess permeation of individual SFs. To characterize its transepithelial transport, different incubation conditions were investigated on Caco-2 monolayer. Transport at 37⬚C, but not at 4⬚C, is indicative of active-transport; while transport occurring at both temperatures suggests a passive diffusion mechanism [17, 20]. The low temperature can significantly decrease the fluidity of cell membrane and cell metabolism including the ATPases activity that is involved in the energy-dependent transport process [21, 22]. The effect of SF 1 on the tight junction of Caco-2 cell monolayer was investigated by measuring the TEER values and cotransport with fluorescent probe LY. Two concentrations of SF 1 were used, 70 M and 190 M. TEER value is temperature-dependent, thus the temperature should be well controlled [17]. In our study, TEER values were constant (around 550 ⍀•cm2 at 37⬚C, 900 ⍀•cm2 at 4⬚C) www.mnf-journal.com
D. Zhu et al.
A
600 550
TEER (
cm2)
500 450 400 350
LY SF1 (70 µM) LY + SF1 (70 µM) SF1 (190 µM) LY + SF1 (190 µM)
300 250 0
2
4
6
18
Time (h) B 1000
900
cm2)
during the incubation with SF 1 at 70 M, as well as its coincubation with LY (Fig. 3, A and B). The used vehicle had no effect on the TEER. This suggests that the integrity of Caco-2 cell monolayers was not impaired during the transport studies. When increasing the concentration to 190 M, the membrane integrity was transiently compromised, as indicated by the significant decrease of TEER values by ࣈ 25% at 37⬚C. This was likely due to the higher amount of vehicle required rather than the SFs themselves. Nevertheless, the cells regained their integrity within 18 h (Fig. 3A). Similar behavior in TEER values were also observed when performed at 4⬚C (Fig. 3B). LY is known to cross Caco-2 cell monolayer exclusively through passive paracellular diffusion [18]. When the Papp of LY is smaller than 1 × 10−6 cm/s, Caco-2 cell monolayers are suitable for the permeation study [18]. In our study, Papp of LY alone in monolayer was 5.4 × 10−8 cm/s. No significant changes for Papp of LY were observed with SF 1 at 70 M both at 4 and 37⬚C (Fig. 3C, calculation in Supporting Information S3). When the concentration of SF 1 was raised to 190 M, the Papp value of LY was slightly increased at 37⬚C, Papp was increased even further by reducing the temperature to 4⬚C, suggesting some toxic effect at high concentration of SF 1. These observations are in well agreement with cytotoxicity and TEER results; nevertheless, although the toxicity was observed, the Papp value of LY was still low ( 1 × 10−6 cm/s, whereas poorly absorbed drugs have Papp values < 1 × 10−7 cm/s [23]. The low Papp value indicates that SF 1 is a poorly absorbed compound. Moreover, SF 1 had similar Papp values at 4⬚C and 37⬚C, and no significant inhibition of permeation of SF 1 at 4⬚C, suggesting that the permeation of SF 1 is mainly through passive diffusion; nevertheless, it remains unknown whether the paracellular or transcellular diffusion occurs in the Caco-2 cell monolayer.
Mol. Nutr. Food Res. 2015, 59, 1182–1189
TEER (
1186
800
700
600 LY + SF1 (70 µM) LY + SF1 (190 µM)
500 0
1
2
3
4
5
6
Time (h)
C LY +SF1 (190 µM) at 4°C
*
LY + SF1 (190 µM) at 37°C
LY + SF1 (70 µM) at 4°C
* LY + SF1 (70 µM) at 37°C
LY + V (70 µM )
LY
0
4
8
12
16
Permeability coefficient of LY (×10 - 8 cm/s)
3.3 Individual SFs The ability of individual SFs to permeate was investigated using Caco-2 cell monolayer. Given that 190 M of SFs showed some cytotoxicity, likely due to the higher amount of vehicle, all additional SFs were tested at 70 M. Monitoring the integrity of the monolayer (through TEER measurements) suggests no significant cytotoxicity or compromise of the C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Figure 3. Effects of apical exposure of cycloartenyl ferulate (SF 1) in different conditions on the tight junction of Caco-2 monolayer: (A) Variation of TEER of Caco-2 monolayer at 37⬚C. (B) Variation of TEER of Caco-2 monolayer at 4⬚C. (C) Permeability coefficient of LY in Caco-2 monolayer. Data are expressed as mean ± SEM (n = 3–6). An asterisk indicates a significant difference between groups (p < 0.05).
www.mnf-journal.com
1187
Mol. Nutr. Food Res. 2015, 59, 1182–1189
A SF1 (70 µM) at 37°C SF1 (70 µM) + LY at 37°C SF1 (70 µM) + LY at 4°C SF1 (190 µM) at 37°C SF1 (190 µM) + LY at 37°C SF1 (190 µM) + LY at 4°C
0.3
0.2
0.1
0.0 0
SF1 SF2 SF3 SF4 SF5 SF6 SF7 SF8
0.3
1
2
3
Time (h)
Cumulative percentage transported (%)
Cumulative percentage transported (%)
A
0.2
0.1
0.0 0
B
1
2
3
Time (h)
B
SF1 (190 µM) + LY at 4°C
SF8
SF1 (190 µM) + LY at 37°C
SF7 SF1 (190 µM) at 37°C
SF6 SF1 (70 µM) + LY at 4°C
SF5 SF1 (70 µM) + LY at 37°C
SF4
SF1 (70 µM) at 37°C
SF3
0
4
8
Permeability coefficient (×10
12 -8
SF2
16
cm/s)
Figure 4. Permeation of cycloartenyl ferulate (SF 1) in different conditions in Caco-2 monolayer: (A) transport percentage of SF 1 (solid line for 70 M, dashed line for 190 M) over time; (B) comparison of the permeability coefficients of SF 1. Data are expressed as mean ± SEM (n = 6–9).
monolayer, being the resistance consistently at 550 ⍀•cm2 (Supporting Information S5). Although their transport ratios were very low, all the individual SFs showed a time-dependent passage across the model (Fig. 5A). Only 0.23% of SF 1 and 0.22% of SF 6 permeated to the receiver chamber, while for SF 8, the permeation ratio was only 0.04%. In this study, mass recovery was consistently around 100% for all individual SFs, which indicates no significant metabolism or degradation of SFs in monolayer. Generally, all of these SFs had very low Papp value (1.69 × 10−8 –1.10 × 10−7 cm/s), suggesting they are poorly absorbed compounds (Fig. 5B; calculation in Supporting Information S5); nevertheless, up to a 6.5-fold difference was observed between the highest and lowest Papp values (SF 1 versus 8). Statistically, Papp of SF 1 was significantly higher than SF 7 and 8, as well as SF 5 was significantly higher than 8. This study confirmed that individual SFs have very low permeation, while still having significant differences in their permeation values between SF species. C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
SF1
0
4
8
12
16
Permeability coefficient (×10 - 8 cm/s)
Figure 5. Permeation of individual SFs (SF 1–8, 70 M) in Caco2 monolayer at 37⬚C: (A) transport percentage of SFs over time; (B) comparison of the permeability coefficients of SFs. Data are expressed as mean ± SEM (n = 6–9). An asterisk indicates a significant difference between groups (p < 0.05).
4
Discussion
Previous in vivo studies suggested the low absorption behavior of SFs. Fujiwara and coworkers reported that after oral administration of radiolabeled SF to rats, only a small fraction of radioactivity transferred to blood, with a peak concentration corresponding to 0.06% of the dose 4 h after administration [9]. From the available clinical human study by Lubinus and coworkers, no intact SF was reported in the blood sample after consumption of SFs enriched yogurt, and only cycloartenol, which was thought as the product from SF 1, was significantly increased by 23.7% compared on day 3 and day 0 [12]. This suggested that metabolism or degradation of SF 1 might have occurred in the gastrointestinal (GI) tract. Furthermore, Umehara and coworkers reported that in healthy volunteers the peak plasma concentration of SFs was 21–107 ng/mL www.mnf-journal.com
1188
D. Zhu et al.
(equivalent to 0.01–0.05% of the total administered dose) after an oral dose of 600 mg, and 112 ng/mL (< 0.3% of the dose) after repetitive oral dose of 100 mg three times a day for 10 days [24]. However, no information was so far available about the intestinal absorption of individual SFs, hence a different behavior between individual SFs cannot be excluded. In this study, we optimized the permeation conditions using a model compound SF 1, to investigate the permeation of an array of naturally occurring SFs. Although very small amount (less than 0.5%) of individual SFs permeated across the Caco-2 cell monolayer, we still observed that some differences of Papp were found between individual SFs, Papp of SF 1 > 7, 1 > 8, and 3 > 7. Examining the structural differences between these pairs of compounds lead us to believe that SF structures may influence their ability to permeate across the monolayer. SF 1 and 7 or 8 differ in both their sterol rings and side chains, SF 3 and 7 only differ in the presence and absence of one carbon at C-24 position. There is some evidence within literature that various sterols have different absorptions in the animal model, which was associated to the variation in their side chain (campesterol > sitosterol > stigmasterol) and saturation of double bond (sitosterol > sitostanol) [25]. In the view of physicochemical properties of eight SFs, we applied semi-empirical quantum mechanics method (AM1) for molecular simulation to explain their Papp differences. The dipole moment and quantitative structureactivity relationship parameters (logP, surface area, volume, hydration energy, etc.) were obtained from the quantum mechanics simulations. However, we could not identify a distinct trend between their physicochemical properties and permeability (Supporting Information S6). Moreover, high mass recovery after permeation in this study suggested there was no significant hydrolysis, metabolism, or degradation of SF by the cells. SFs were initially hypothesized to be hydrolysed to free ferulic acid and phytosterol, and could subsequently be absorbed [26]. Indeed, some in vitro studies have shown that SFs are substrate for cholesteryl esterase from animal pancreas and microorganism, and cholesteryl esterase has strong preference for desmethylsteryl ferulate [8, 27, 28]. However, Huang found that the decrease in SFs concentration was not accompanied by either an increase in free phytosterols or in ferulic acid in in vitro cell model, indicating that hydrolysis of SF is still unclear [11]. Moreover, the hydrolysis is still possibly present by pancreatic and microbial cholesteryl esterase in vivo [12]. The Caco-2 cell model lacks gut microorganisms that may secrete cholesterol esterase; the significant hydrolysis of SF is not expected. Without additional analysis we cannot entirely exclude the possibility of SF degradation or hydrolysis occurring to a minor degree. Furthermore, the similar permeation of SF 1 both at 4⬚C and 37⬚C indicated that the transport mechanism functions primarily through passive diffusion. Nevertheless, it remains to be determined whether it occurs through paracellular or transcellular diffusion. The similar behaviors in permeation between SFs (SF 2–8) to
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Nutr. Food Res. 2015, 59, 1182–1189
SF 1 (e.g. the similar TEER recording or low Papp values) lead us to conclude that they also permeate through an analogous passive diffusion mechanism. Still, active transport cannot be fully excluded within this study. In our study, the low Papp values resulted in extremely low levels of transported SFs and therefore in an analytical limitation. This hinders further analysis of potential hydrolytic products of SFs, as well as the analysis of a possible active transport. SFs are primarily found in the bran of certain cereal grains and seeds. Their biosynthetic pathway as well as location within the plant cells remains unknown. Their function in plants may be related to protection against environmental stress [29], antioxidant property [30], and regulation of microbial activity in the grains [1]. Dietary intake of SFs usually comes from cereal grains like rice, wheat and corn. However, according to our previous study, the bioaccessibility of SFs was found to be almost negligible (0.1%) from a grain matrix [2], which indicates that SFs do not need to be highly absorbable from the complex grain matrix to provide their functional effect. Furthermore, the low absorption behavior, confirmed in this study, suggests that the cholesterol-lowering and antioxidant activity-related health benefits associated with SFs most likely occur in the gut independently from absorption. For instance, a recent work by M¨akynen and coworkers showed that SFs (␥-oryzanol) significantly inhibited incorporation of cholesterol into synthetic micelles, and also SFs had ability to significantly decrease the apical uptake of cholesterol into Caco-2 cells [31]. Moreover, many of the poorly absorbed dietary antioxidants indeed play an important role in protecting the GI tract from oxidative damage without necessarily being absorbed [32]. Their concentration can be much higher in the lumen of GI tract than are ever achieved in plasma or tissues, and their antioxidant action in the GI tract can delay the development of inflammation and related disease (e.g. cancer) in stomach or colon [33]. SFs showed the potential to ameliorate colonic inflammation in colitis mice model by inhibiting NF-B activity and decreasing proinflammatory cytokines and COX-2 levels due to their antioxidant effects [34]. Furthermore, they may act as antioxidants against reactive oxygen species produced by intestinal microflora. Little is known about the bioactivities of SFs in the digesta, an area that remains to be explored. The results of this work confirm that all the individual SFs share the same behavior: they are poorly absorbed compounds and their health benefits most likely occur in the gut independently from absorption. This work was supported by ETH Zurich and doctoral scholarship from Chinese Scholarship Council (CSC). D. Brambilla gratefully acknowledges support from the ETH Zurich Postdoctoral Fellowship (2012-01). We thank Dr. Antoni S´anchez-Ferrer (Laboratory of Food and Soft Materials, ETH Zurich, Switzerland) for providing stigmasteryl and cholesteryl ferulate synthesis and computer simulation, and thank Dr. Anja Rahn for English editing. The authors have declared no conflict of interest.
www.mnf-journal.com
Mol. Nutr. Food Res. 2015, 59, 1182–1189
5
References
[1] Seitz, L. M., Stanol and sterol esters of ferulic and p-coumaric acids in wheat, corn, rye, and triticale. J. Agric. Food Chem. 1989, 37, 662–667. ¨ [2] Mandak, E., Nystrom, L., The effect of in vitro digestion on steryl ferulates from rice (Oryza sativa L.) and other grains. J. Agric. Food Chem. 2012, 60, 6123–6130.
1189 [18] Yu, H., Huang, Q., Investigation of the absorption mechanism of solubilized curcumin using Caco-2 cell monolayers. J. Agric. Food Chem. 2011, 59, 9120–9126. [19] Fang, N., Yu, S., Badger, T. M., Characterization of triterpene alcohol and sterol ferulates in rice bran using LC-MS/MS. J. Agric. Food Chem. 2003, 51, 3260–3267.
[3] Ghatak, S. B., Panchal, S. J., Gamma-oryzanol-A multipurpose steryl ferulate. Curr. Nutr. Food Sci. 2011, 7, 10–20.
´ [20] Rieux, A. d., Ragnarsson, E. G. E., Gullberg, E., Preat, V. et al., Transport of nanoparticles across an in vitro model of the human intestinal follicle associated epithelium. Eur. J. Pharm. Sci. 2005, 25, 455–465.
[4] Cicero, A., Derosa, G., Rice bran and its main components: potential role in the management of coronary risk factors. Curr. Top. Nutraceutical Res. 2005, 3, 29–46.
[21] Los, D. A., Murata, N., Membrane fluidity and its roles in the perception of environmental signals. Biochim. Biophys. Acta 2004, 1666, 142–157.
¨ [5] Nystrom, L., in: Xu, Z., Howard, L. R. (Eds.), Analysis of Antioxidant-Rich Phytochemicals, Wiley-Blackwell, Oxford, UK 2012, pp. 313–351.
[22] Holtzman, E., in: Holtzman, E. (Ed.), Lysosomes, Springer US 1989, pp. 25–92.
[6] Bergman, C. J., Xu, Z., Genotype and environment effects on tocopherol, tocotrienol, and ␥-oryzanol contents of southern U.S. rice. Cereal Chem. 2003, 80, 446–449. ¨ ¨ [7] Nystrom, L., Makinen, M., Lampi, A. M., Piironen, V., Antioxidant activity of steryl ferulate extracts from rye and wheat bran. J. Agric. Food Chem. 2005, 53, 2503–2510. ¨ [8] Nystrom, L., Moreau, R. A., Lampi, A.-M., Hicks, K. B., Piironen, V., Enzymatic hydrolysis of steryl ferulates and steryl glycosides. Eur. Food Res. Technol. 2008, 227, 727–733. [9] Fujiwara, S., Sakurai, S., Sugimoto, I., Awata, N., Absorption and metabolism of gamma-oryzanol in rats. Chem. Pharm. Bull. 1983, 31, 645–652. [10] Fujiwara, S., Sakurai, S., Noumi, K., Sugimoto, I., Awata, N., Metabolism of gamma-oryzanol in rabbit. Yakugaku zasshi 1980, 100, 1011–1018. [11] Huang, C. J., Potential functionality and digestibility of oryzanol as determined using in vitro cell culture models. PhD Thesis. Louisiana State University 2003. [12] Lubinus, T., Barnsteiner, A., Skurk, T., Hauner, H., Engel, K. H., Fate of dietary phytosteryl/-stanyl esters: analysis of individual intact esters in human feces. Eur. J. Nutr. 2013, 52, 997–1013. [13] Carey, M. C., Small, D. M., Bliss, C. M., Lipid digestion and absorption. Annu. Rev. Physiol. 1983, 45, 651–677. [14] Hubatsch, I., Ragnarsson, E. G., Artursson, P., Determination of drug permeability and prediction of drug absorption in Caco-2 monolayers. Nat. Protoc. 2007, 2, 2111–2119. ¨ [15] Mandak, E., Zhu, D., Godany, T. A., Nystrom, L., Fourier transform infrared spectroscopy and Raman spectroscopy as tools for identification of steryl ferulates. J. Agric. Food Chem. 2013, 61, 2446–2452. [16] Hamada, T., Goto, H., Yamahira, T., Sugawara, T. et al., Solubility in and affinity for the bile salt micelle of plant sterols are important determinants of their intestinal absorption in rats. Lipids 2006, 41, 551–556. ˚ o, ¨ J., Taipalensuu, J., Ocklind, G., Artursson, [17] Tavelin, S., Grasj P., in: Wise, C. (Ed.), Epithelial Cell Culture Protocols, Humana Press, Totowa 2002, pp. 233–272.
C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
[23] Artursson, P., Palm, K., Luthman, K., Caco-2 monolayers in experimental and theoretical predictions of drug transport. Adv. Drug Deliv. Rev. 2001, 46, 27–43. [24] Umehara, K., Shimokawa, Y., Miyamoto, G., Effect of gamma-oryzanol on cytochrome P450 activities in human liver microsomes. Biol. Pharm. Bull. 2004, 27, 1151–1153. [25] Ling, W. H., Jones, P. J., Dietary phytosterols: a review of metabolism, benefits and side effects. Life Sci. 1995, 57, 195– 206. [26] Berger, A., Rein, D., Schafer, A., Monnard, I. et al., Similar cholesterol-lowering properties of rice bran oil, with varied gamma-oryzanol, in mildly hypercholesterolemic men. Eur. J. Nutr. 2005, 44, 163–173. [27] Miller, A., Majauskaite, L., Engel, K.-H., Enzyme-catalyzed hydrolysis of ␥-oryzanol. Eur. Food Res. Technol. 2004, 218, 349–354. [28] Moreau, R. A., Hicks, K. B., The in vitro hydrolysis of phytosterol conjugates in food matrices by mammalian digestive enzymes. Lipids 2004, 39, 769–776. [29] Britz, S. J., Prasad, P. V., Moreau, R. A., Allen, L. H. J. et al., Influence of growth temperature on the amounts of tocopherols, tocotrienols, and gamma-oryzanol in brown rice. J. Agric. Food Chem. 2007, 55, 7559–7565. [30] Nantiyakul, N., Furse, S., Fisk, I., Foster, T. et al., Phytochemical composition of Oryza sativa (rice) bran oil bodies in crude and purified Isolates. J. Amer. Oil Chem. Soc. 2012, 89, 1867– 1872. ¨ [31] Makynen, K., Chitchumroonchokchai, C., Adisakwattana, S., Failla, M., Ariyapitipun, T., Effect of gamma-oryzanol on the bioaccessibility and synthesis of cholesterol. Eur. Rev. Med. Pharmacol. Sci. 2012, 16, 49–56. [32] Stahl, W., vanden Berg, H., Arthur, J., Bast, A. et al., Bioavailability and metabolism. Mol. Aspects Med. 2002, 23, 39–100. [33] Halliwell, B., Zhao, K., Whiteman, M., The gastrointestinal tract: a major site of antioxidant action? Free Radic. Res. 2000, 33, 819–830. [34] Islam, M. S., Murata, T., Fujisawa, M., Nagasaka, R. et al., Anti-inflammatory effects of phytosteryl ferulates in colitis induced by dextran sulphate sodium in mice. Br. J. Pharmacol. 2008, 154, 812–824.
www.mnf-journal.com
